Classification of drug-related problems with new prescriptions using a modified PCNE classification system
Objectives To explore and classify drug-related problems (DRPs) with new prescriptions detected in community pharmacies using a modified PCNE (Pharmaceutical Care Network Europe) classification system. Setting Sixty-four Swiss community pharmacies offering internships for pharmacy students. Main outcome measures Occurrence, nature and pharmacist’s management of DRPs. Methods Fifth-year pharmacy students collected consecutively hospital discharge and primary care prescriptions. After training, they documented clinical and technical DRPs, causes and interventions. Results Prescriptions of 616 patients (43.0% discharged from hospital) were analysed. The patients’ median age was 56 years and they received a median of 3 (range 2–19) different drugs. In 121 (19.6%) prescriptions 141 clinical DRPs were detected. The most frequent clinical DRPs were potential drug–drug interactions (DDIs) (37.6%), drug choice (24.8%) and drug use problems (15.6%). These clinical DRPs led to a total of 299 interventions. There were 222 prescriptions (36.0%) that showed 278 technical DRPs, resulting in a total of 417 interventions. Most frequent technical DRPs were missing or unclear package size or therapy duration (32.7%) and missing or unclear dosing/application instructions (30.9%). Most DRPs (75.4%) could be managed by the pharmacist alone. The number of prescribed drugs was the main factor with an influence on the frequency of clinical and technical DRPs. Conclusion Clinical and technical DRPs are frequently observed in primary care as well as in hospital discharge prescriptions. The modified PCNE classification system, especially the amendment with a technical DRP category, proved to be useful and allowed the classification of all DRPs. Neither the setting (hospital discharge vs. primary care) nor the quality of electronically printed prescriptions, but only the number of prescribed drugs influenced the occurrence of clinical or technical DRPs.
KeywordsClassification system Community pharmacy Drug-related problems Hospital discharge PCNE Pharmaceutical care Primary care Switzerland
Impact of findings on practice
In Switzerland, half of all new prescriptions show a drug-related problem; two-thirds of those are technical drug-related problems.
The PCNE classification V5.1 needs to be amended, to include technical drug-related problems.
The occurrence of clinical or technical drug-related problems is only influenced by the number of prescribed drugs. It is not relevant if the prescriptions were electronically printed.
Many studies have shown drug-related problems (DRPs) to be very common in primary care and in hospital settings [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16]. In both settings there is evidence that pharmacists’ interventions can reduce the occurrence of DRPs [6, 9, 14, 15, 17].
Receiving a newly prescribed drug is most likely an extraordinary situation for a patient who was recently informed about a diagnosis or at least was confronted with a new drug in his regimen. Because the risk of DRPs may be increased on the initiation of new drug treatments or changes within an established drug-treatment plan, a thorough consultation with a pharmacist is required to consider the new medication and, in particular, to make an accurate check for DRPs to achieve desired health outcomes [4, 18]. Thus, patients with at least one newly prescribed drug represent a relevant population to study DRPs and especially for studying the applicability of a comprehensive classification system which includes technical DRPs.
Usually, patients discharged from hospital also have new drugs prescribed and are at increased risk of experiencing DRPs . Therefore, we decided to focus on both prescriptions from hospital discharge and new primary care prescriptions, expecting that there are differences between these two settings (e.g. pattern and number of DRPs) as well as differences between electronically printed and handwritten prescriptions. To our knowledge, no previous study focused on new primary care and hospital discharge prescriptions processed in Swiss community pharmacies in everyday life.
In different studies, different detection rates of DRPs were found because of variations in the methods used to identify DRPs, the classification systems and the inclusion criteria. There is no accepted standard tool for classification and documentation of DRPs, for the primary care or hospital (discharge) settings. The applicability of different systems is not yet clear, and many studies conclude that further studies are needed, with the aim to provide a tool that allows a complete classification of all DRPs that arise during prescription processing in community pharmacies.
From different pilot studies we had experience in the use of the PCNE classification (PCNE Classification V 5.01) . This system attributes at least four items to each observation: (a) coding for the problem itself, (b) the actual or suspected cause(s) of the problem, (c) the intervention(s) required to resolve the DRP and (d) its outcome. We recognised as a probably important deficiency the lack of possibilities to classify technical DRPs, which arise frequently from prescription processing in community pharmacies (e.g. missing or unclear specification of drug dosage form, drug dosage and dosage regimen). In a very recent study on pharmacists’ interventions during the prescription dispensing process of new prescriptions for acute respiratory tract infections, we observed a need for addition of specifications in 11.6% of all prescription items . Therefore, we planned to amend the existing PCNE classification accordingly.
The first objective of this study was to explore the occurrence, nature and pharmacist’s management of drug-related problems (DRPs) detected in community pharmacies using the modified PCNE classification system in new prescriptions. The second aim was to analyse possible differences between new primary care and hospital discharge prescriptions as well as differences between electronically printed and handwritten prescriptions. The third aim was to evaluate the applicability of the modified classification system.
Setting and participants
This prospective observational study was conducted from January to April 2007 in 64 Swiss community pharmacies offering internships for fifth-year pharmacy students. Prescriptions of patients aged 18 years or older were eligible for the study if they comprised at least two prescribed drugs and if at least one drug was new (a new drug was defined as a new drug, drug form or dosage or a generic drug but not a new package size). In Switzerland not every prescription contains a new drug because the health care system allows physicians to issue repeat prescriptions and thus enable patients to get their medications up to 12 months without further visit of the physician. Prescriptions issued by primary care physicians, physicians from ambulatory care centres or outpatient clinics of hospitals were classified as ‘primary care prescriptions’; ‘hospital discharge prescription’ was defined as a prescription issued by a hospital after the patient had been admitted for least one night.
Sixty-four fifth-year pharmacy students had the mandatory mission from the university to collect five hospital discharge and five primary care prescriptions in a consecutive way during their pharmacy internships, leading to a convenient sample of prescriptions with at least one new medication. They were free to select one day to do so within the designated study period. While serving clients as usual they only had to identify if they were in possession of a prescription with a new drug and to check for inclusion criteria regardless whether they identified any problem or not. Then they were to check for DRPs with all the information available in a standard pharmacy, including the patient’s medication history. The software used in all Swiss community pharmacies performs computerized screening and alerts the attending employee of potential drug-drug interactions (DDIs). Whether a potential DDI was documented as a DRP or not depended on the student and the pharmacist: if the DDI was seen as a problem for the patient they had to complete a classification form.
Thereafter, prescription validation was performed as usual by a pharmacist. Accordingly, the pharmacist or the student could detect a DRP or a problem that arose during patient counselling or processing the prescription. After they had included the first prescription they were obliged to collect the subsequent nine prescriptions fulfilling the inclusion criteria leading to five hospital and five primary care prescriptions. This consecutive collecting enabled the estimation of the incidence of DRPs with new prescriptions and the compilation of a randomized, convenient sample to reflect the daily life setting in Swiss community pharmacies. For each prescription at least one data sheet had to be filled out and if multiple problems were detected in the same prescription for each problem a new data sheet was used. To each problem multiple causes and interventions could be attributed. Students were instructed to document each case shortly after dispensing the prescribed drugs. We retrieved the number of prescribed drugs per prescription from copies of prescriptions and verified plausibility of students’ documentation. Students anonymised all documents before delivery to the study centre.
Students were trained during a lecture at the university. Three case studies were used to test their performance in documenting problems, causes and interventions resulting in a proportion of 91.9% of correct coding. In addition, ongoing support was assured by the study team being in contact by e-mail or phone.
Classification of DRPs
We used the PCNE classification system for DRPs (version 5.01) . This system distinguishes four dimensions: problems, causes, interventions and the corresponding outcome.
The six main categories of clinical DRPs are: (P1) adverse reaction(s), (P2) drug choice problem, (P3) dosing problem, (P4) drug use problem, (P5) interactions and (P6) other. We added a seventh category (P7) ‘technical DRP’ with the aim to be able to classify and to distinguish between “clinical DRPs” (P1–P6) and ‘technical DRPs’. A technical DRP is related to prescription quality and impedes to unambiguously dispense a drug in the correct dose, dosage form and package size (e.g. unreadable prescription, missing specifications). Accordingly, we created a seventh category of causes: C7 prescription quality (e.g. unreadable prescription, missing or unclear drug form, dosage, package size or therapy length). The pre-existing categories of causes were: (C1) drug/dose selection, (C2) drug use process, (C3) information, (C4) patient/psychological, (C5) (pharmacy) logistics and (C6) other. Categories in the intervention domain remained unchanged: (I0) ‘no intervention’, (I1) at prescriber level, (I2) at patient level, (I3) at drug level, (I4) other. Unchanged outcome categories were: (O0) outcome unknown, (O1) problem solved, (O2) problem partially solved, and (O3) problem not solved.
Added items to the original PCNE classification system V5.01
C7 prescription quality
P4.3 wrong or improper application method of drug (e.g. tablets without break score cut into halves)
C7.1 unreadable prescription
I2.5 patient asked for further information
P4.4 wrong or improper time of application or intake
C7.2 missing or unclear drug name, though readable
I2.6 more information retrieved from patient’s history
P7.1 technical DRP
C7.3 missing or unclear drug form if multiple forms are available
I3.7 adaptation of amount of drug dispensed
C7.4 missing or unclear drug potency if multiple forms are available
I4.3 information retrieved from literature
C7.5 missing or unclear package size or therapy length
I4.4 information retrieved from toxicology or drug information centre
C7.6 missing or unclear dosage or application instruction
C7.7 missing prescription of necessary applications aids
Users’ opinion on the usability of the tool
To evaluate usability of and satisfaction with the modified classification system, an 8-item questionnaire used in a prior published study  was administered to the students who used this tool. They were asked about the extent to which they agree or disagree (five-point scale) with statements regarding the usability and usefulness, satisfaction and their willingness to use the tool in their practice in the future.
All returned data sheets were processed with the automated forms processing software TeleForm® version 7.0 from Cardiff Software Inc., Vista, USA. To avoid potential errors, all numeric and letter recognitions were verified visually on data sheets and on screen.
Basic study characteristics
Prescriptions (=patients)—n (%)
Age—mean years (SD, median) (IQR, range)
53.3 (19.0;53) (28.5; 18–98)
59.3 (18.6; 62) (27.3; 19–94)
55.9 (19.0; 56) (32; 18-98)
Electronic prescriptions—n (% of prescriptions)
Total number of prescribed drugs—n (%)
Number of drugs per prescription—median (IQR, range)
2 (1; 2–10)
4 (5; 2–19)
3 (2; 2–19)
Clinical DRPs—n (% of prescriptions)
Technical DRPs—n (% of prescriptions)
Prescriptions with clinical DRPs—n (% of prescriptions)
Prescriptions with technical DRPs—n (% of prescriptions)
Prescriptions without any problem—n (% of prescriptions)
During prescription processing in the community pharmacies, 419 clinical and technical DRPs (or both) were detected in 329 prescriptions. In relation to the total number of prescribed drugs (n = 2309) we found 6.1% (n = 141) clinical and 12.0% (n = 278) technical DRPs. All problems could be classified with the modified PCNE classification system.
Multiple clinical problems were present in 12 prescriptions, multiple technical DRPs in 49 prescriptions and 14 prescriptions showed both clinical and technical DRPs.
Clinical DRPs (n = 141) in 121 prescriptions, classified according to PCNE classification V5.1
n (% of 141)
n (% of 141)
n (% of 141)
Side effect suffered (non-allergic)
Side effect suffered (allergic)
Toxic effects suffered
Drug choice problem
Inappropriate drug (not most appropriate for indication)
Inappropriate drug form (not most appropriate for indication)
Inappropriate duplication of therapeutic group or active ingredient
Contra-indication for drug (incl. pregnancy/breast feeding)
No clear indication for drug use
No drug prescribed but clear indication
Drug dose too low or dosage regimen too long
Drug dose too high or dosage regimen too frequent
Duration of treatment too short
Duration of treatment too long
Drug use problem
Drug not taken/administered at all
Wrong drug taken/administered
Wrong/not appropriate drug use/application
Wrong/not appropriate time of use/application
Patient dissatisfied with therapy despite taking drug(s)
Insufficient awareness of health and diseases (possibly leading to future problems)
Unclear complaints, further clarification necessary
Therapy failure (reason unknown)
Technical DRPs (n = 278) in 222 prescriptions
n (% of 278)
n (% of 278)
n (% of 278)
Missing prescription of necessary application aids
Missing/unclear drug name though legible
Missing/unclear drug form, if several available
Missing/unclear drug potency, if several available
Missing/unclear dosing/application instruction
Missing/unclear package size and/or therapy length
The main contributing factor for presentation of clinical and technical DRPs was the number of drugs per prescription. Other variables had no significant influence in our dataset (univariate analysis of variance, analysis of discriminance, logistic regression and correlation).
The probability for a clinical DRP was similar in prescriptions between two and five drugs; if >5 drugs were prescribed then the probability increased (see Fig. 1). The number of drugs is correlated with the chance of a clinical DRP (r = 0.220; P < 0.001; bivariate correlation, Spearman) and with ‘any problems’ (r = 0.132; P = 0.001; bivariate correlation, Spearman) but not with technical DRPs (r = 0.012; P = 0.775; bivariate correlation, Spearman).
The mean number of prescribed drugs was significantly higher for prescriptions from hospital discharge than from primary care (5.1 vs. 2.7; P < 0.001; t test). This higher number of prescribed drugs in hospital discharge prescriptions caused more clinical but not more technical DRPs. To analyse the influence of the setting (primary care and hospital discharge), we only considered prescriptions with ≤5 prescribed drugs because primary care prescriptions rarely included >5 prescribed drugs. We observed significantly more technical DRPs among prescriptions from primary care (P = 0.043; chi-square; 95%; 2-tailed) but no influence of the setting for any problem (P = 0.080; chi-square; 95%; 2-tailed) or clinical DRPs (P = 0.924; chi-square; 95%; 2-tailed). There was no association between gender and problems. Because the relationship between the number of prescribed drugs and technical DRPs is not linear and we do not have sufficient data we could not statistically correct for this effect.
Top 10 of the most frequently reported causes (n = 166) induced by clinical DRPs (n = 141), classified according to PCNE classification V5.1
n (% of 166)
n (% of 166)
n (% of 166)
Inappropriate drug selection
Inappropriate dosage selection
Pharmacokinetic problems, incl. ageing/deterioration in organ function and interaction
Drug use process
Inappropriate timing of administration and/or dosing intervals
Patient unable to use drug (form) as directed
Instructions for use/taking not known
Patient unaware of reason for drug treatment
Burden of therapy
Prescribed drug not available
Top 10 of the most frequently reported interventions (n = 716) induced by clinical or technical DRPs, classified according to PCNE classification V5.1
n (% of 716)
n (% of 716)
n (% of 716)
n (% of 716)
n (% of 716)
At prescriber level
Prescriber informed only
Asked prescriber for further information
Intervention proposed, approved by prescriber
At patient/carer level
Patient (medication) counselling
Patient asked for further information
Further information retrieved from patient history
At drug level
Instructions for use changed
Amount of drug to dispense changed
To manage all problems a total of 81 (13.1% out of 616 prescriptions) direct contacts to the prescriber (call/conversation) were necessary: 42 (6.8% out of 616) prescriptions with one or more clinical DRPs and 46 (7.5% out of 616) prescriptions with one or more technical DRPs led to a direct contact with the prescriber (seven prescriptions with both a clinical DRP and a technical DRP). One hundred and seventy-five (63.0% of 278) of technical DRPs could be solved by the pharmacy together with the patient, but only 56 (39.7% out of 141) of clinical DRPs were solved in cooperation with the patient.
The 299 interventions induced by clinical DRPs involved 178 drugs (in 48 cases two drugs were related to one problem, e.g. interaction). Cardiovascular drugs (C; 21.9%, n = 39) were most often involved, followed by nervous system (N; 21.3%, n = 38) and musculo-skeletal system drugs (M; 10.1%, n = 18).
Users’ opinion (n = 64) of the tool
Mean score ± SD
Agree or strongly agree
Disagree or strongly disagree
1. The classification system was comprehensive and included all drug-related problems I identified.
3.4 ± 1.1
2. I did not have problems finding out the proper classification of drug-related problems I identified.
2.9 ± 1.0
3. The classification system was easy to use and practical.
3.1 ± 0.9
4. I will use the classification system in my practice in the future.
2.3 ± 1.3
5. In general I am satisfied with the classification system.
3.3 ± 1.0
6. The expenditure of time to classify the problems was adequate.
3.7 ± 1.0
7. I think it is important to have an opportunity to classify the effort of pharmacies.
4.4 ± 0.7
8. The PCNE classification would be a good tool to document the activities of pharmacies.
3.1 ± 1.1
This study examined the frequency, nature and pharmacist’s management of DRPs with primary care and hospital discharge prescriptions which comprised at least one new drug. We found a high occurrence of clinical (19.6%) and, in particular, of technical DRPs (36.0%). More than 50% of all prescriptions showed a clinical or a technical DRP or both. Compared to other studies our numbers are quite high but in this study we set out to use prescriptions which we considered likely to have a high prevalence, i.e. newly started drugs and prescriptions with at least two drugs. In 2007 Hammerlein found only 0.93 DRPs per 100 patients and 1.16 DRPs per 100 prescriptions . Westerlund conducted a survey on DRPs in prescription-only medicines in 1999 and found 2.8 DRPs per 100 patient contacts . A recent Swiss study detected 0.74 clinical DRPs per 100 prescriptions with the need of an intervention and 1.9 technical DRPs per 100 prescriptions . These figures illustrate the problem that such comparisons are hampered by different settings, measurements methods and classification systems. Paulino  found in a study in six European countries even more DRPs than we did (103.7 DRPs per 100 patients discharged from hospital; 63.7% out of all patients had at least one DRP). Uncertainty or lack of knowledge about the aim of the drug (29.5%), side effects (23.3%) and practical problems (7.6%) were the most common DRPs. However, only few DDIs (4.0%) were detected, probably because most of them did not have access to patient medication histories. In Switzerland every pharmacy is equipped with software that includes an automatic DDI screening system and participants in our study probably indicated every possible potential interaction, even if not clinical relevant. Therefore, the most frequently reported clinical DRPs in our study were DDIs (37.6% of all clinical DRPs; 7.0% of all patients). In the German study  DDIs were also most frequent but the fraction was only 8.5% of all DRPs. In another primary care study DDIs were the third most common problem with 3.2%, and in a study performed in hospital setting they observed 17.0% DDIs [8, 9]. Paulino et al. found 60.0% of patients after hospital discharge with a potential interaction .
However, the comparison of the study’s findings with those of other studies is difficult due to differences in setting and aim [1, 4, 5, 8, 9, 22]. Furthermore, the frequency of detected problems can be influenced by the systematic screening under study conditions, which differs significantly from daily practice (Hawthorne effect ). In our study, students and pharmacists probably had a more positive attitude towards the provision of pharmaceutical care (e.g. check for DRPs) and they probably detected most possible problems. This may explain our high detection rate of 22.9. If DDIs are as frequent as found in this or other studies more emphasis should be placed on best management of potential DDIs in community pharmacies. However, if we omit all interactions we still found a detection rate of 14.3.
Using fifth-year pharmacy students for data collection during their internship and collecting primary care as well as hospital discharge prescriptions in a continuous way, enabled us to compile a randomized, convenient sample and to estimate the prevalence of DRPs with new prescriptions. We had instructed the students to perform prescription processing as usual. Thus, every prescription was checked by a pharmacist before dispensing and we were not depending on inferior competencies of students. Factors associated with clinical and technical DRPs were tested with several statistical methods and the main factor influencing the presence of any clinical problems was the number of drugs. Older people showed both more clinical and technical DRPs, but this was to be expected because the number of drugs increased with age. In addition, we expected the setting to have an influence on the occurrence of DRPs but different statistical tests did not show any relationship. More clinical DRPs were found in hospital discharge than in primary care prescriptions but the mean number of prescribed drugs was higher on these prescriptions, which explains the higher frequency of problems. Surprisingly, more technical DRPs were found in primary care prescriptions with fewer prescribed drugs than in hospital discharge prescriptions.
Most problems were caused by drug selection and/or drug dose selection and many problems were solved after discussion with patients or family members. More than two-thirds of technical but less than half the clinical DRPs could be solved by the pharmacy together with the patient. Pharmacists should therefore actively involve patients and relatives in the screening and solving process for DRPs. Almost two-thirds of all interventions and 50% of all contacts with the prescriber were related to technical DRPs which shows their importance and frequency in daily life practice in community pharmacies. Therefore, inclusion of technical DRPs into the classification is strongly recommended. Looking at the intervention part, pharmacists indicate the importance of the detected clinical DRPs, in particular activities at prescriber and drug level. The high frequency of interventions is related to the high detection rates of DRPs.
We amended in this study the original PCNE system with 80 items by addition of 15 items (incl. technical DRPs as one new main category) resulting in a total of 95 items. This new problem category was specified with seven items to receive enough information. As suggested by Lampert  and Allenet , we further added two items in the category ‘drug use problem’ to categorise wrong or improper use of a drug and wrong or improper time of applying a drug. A previous study showed a documentation rate of 97.8% in a hospital setting without any modification of the classification system but they reported a lack of certain items for in-patients . Therefore, an important result of this study is the fact that our modified system allowed a complete classification of all problems. In contrast, a study  with the PI-Doc system with 72 items was amended with 27 items but 362 out of 10,427 (3.5%) cases still could not be classified. Our study found twice as many technical DRPs as clinical ones. Most of the documentation systems do not separate technical from clinical DRPs and give them the same level of importance, or they have integrated technical DRPs in ‘prescribing errors’ or ‘other’ [1, 19, 25, 26, 27]. Often there is only a definition for clinical but not for technical DRPs. The German study of 2007 reported 8.3% (out of 10,427) of technical DRPs and difficulties to classify technical DRPs in over 50% of the cases . Our results strongly support the inclusion of a new problem category ‘technical DRPs’ with enough further possibilities to specify them. Furthermore, different classification systems should feature at least the same main categories with the possibility to change single items to be able to compare results from different studies.
Our study has strengths and limitations. We selected prescriptions from primary care during prescription processing in a daily life setting. We could ascertain the reported problems with the copies of the prescriptions and we could retrospectively differentiate between handwritten and electronically written prescriptions. All problems could be coded with our modified PCNE classification system. However, we did not assess the level of severity and the clinical or economic impact of clinical and technical DRPs. Furthermore, we cannot assure that the prescriptions selected were consecutive ones and the diversity of persons who documented clinical and technical DRPs could lead to variability in coding for the same problem, even though the participant observers were trained with specific and detailed instructions. In our study the PCNE system showed to be easy to use for a majority of the users and took little time, as it has been created for the documentation of DRPs in the community pharmacy setting. However, one quarter of students evaluated the PCNE tool as not comprehensive although they had classified all problems. Our students rated very critically all questions if comparing with the study from which the questions were retrieved .
Our study reflects only the situation in Switzerland. A recent multinational study by Mc Elnay et al. investigated the extent to which pharmaceutical care has already been implemented into daily community pharmacy practice across Europe (Mc Elnay J, Hughes C. Provision of pharmaceutical care by community pharmacists: a comparison across Europe). The behavioural pharmaceutical care scale was used and the mean of all 13 countries was 72.1 ± 8.5 (mean ± SD) with a range of mean scores from 50.6 to 83.5 (possible score: 15–160) reflecting a considerable variation between countries. Switzerland had a score of 73.2 ± 18.7 (mean ± SD) indicating that Swiss pharmacists behave most probably not very differently from their counterparts in other European countries. Furthermore, this study revealed an important lack of documentation activities throughout Europe: only 25.2% of all possible documentation possibilities were used. Thus, a documentation and classification system is urgently needed. Intensive training can increase the number of interventions without increasing the time spent on documentation [28, 29]. As Lampert  reported, the PCNE system with its classification of each DRP on four different levels gives enough detail to allow qualitative and even economic analyses.
In the delivery process of new prescribed medications, clinical and technical DRPs are frequently observed in new primary care and in hospital discharge prescriptions. Neither the setting (hospital discharge vs. primary care) nor the quality of electronically printed prescriptions, but only the number of prescribed drugs influenced the occurrence of clinical or technical DRPs in this study. Most DRPs could be managed by the pharmacist alone or after discussion with the patient but without any contact to the prescriber. Therefore, the management of DRPs in community pharmacies is a very important activity which should be explored more intensively in further studies. The modified PCNE classification system, especially the amendment with a technical DRP category, proved to be useful and allowed the classification of all DRPs, but still rather complicated to apply in pharmacy practice.
We would like to thank the owners of the 64 participating pharmacies and Michael Mittag for the support in statistical analysis.
No funding was provided for this Ph.D. student research.
Conflicts of Interest
- 1.Westerlund LT, Almarsdottir AB, Melander A. Drug-related problems and pharmacy interventions in community practice. Int J Pharm Pract. 1999;7(11):40–50.Google Scholar
- 16.Shalansky S, Nakagawa R, Wee A. Drug-related problems identified and resolved using pharmaceutical care versus traditional clinical monitoring. Can J Hosp Pharm. 1996;49(6):7.Google Scholar
- 19.PCNE. The PCNE Classification V 5.01 2006 29 May 2006 [cited 20 Sep 2009]. Available from: http://www.pcne.org/dokumenter/DRP/PCNE%20classification%20V5.01.pdf.
- 29.Currie JD, Chrischilles EA, Kuehl AK, Buser RA. Effect of a training program on community pharmacists’ detection of and intervention in drug-related problems. J Am Pharm Assoc (Wash). 1997;NS37(2):182–91.Google Scholar